image.png

In [1]:
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
In [2]:
from sklearn.datasets import load_iris
iris = load_iris(as_frame=True)
type(iris)
Out[2]:
sklearn.utils.Bunch
In [3]:
df=iris.frame
type(df)
Out[3]:
pandas.core.frame.DataFrame
In [4]:
#iris.feature_names
#df = pd.DataFrame(iris.data,columns=iris.feature_names)
In [5]:
df.head()
Out[5]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) target
0 5.1 3.5 1.4 0.2 0
1 4.9 3.0 1.4 0.2 0
2 4.7 3.2 1.3 0.2 0
3 4.6 3.1 1.5 0.2 0
4 5.0 3.6 1.4 0.2 0
In [6]:
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 150 entries, 0 to 149
Data columns (total 5 columns):
 #   Column             Non-Null Count  Dtype  
---  ------             --------------  -----  
 0   sepal length (cm)  150 non-null    float64
 1   sepal width (cm)   150 non-null    float64
 2   petal length (cm)  150 non-null    float64
 3   petal width (cm)   150 non-null    float64
 4   target             150 non-null    int32  
dtypes: float64(4), int32(1)
memory usage: 5.4 KB
In [7]:
df['species'] = df['target'].replace(to_replace=[0,1,2], value=['Iris-setosa','Iris Versicolour','Iris Virginica'])
df.head()
Out[7]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) target species
0 5.1 3.5 1.4 0.2 0 Iris-setosa
1 4.9 3.0 1.4 0.2 0 Iris-setosa
2 4.7 3.2 1.3 0.2 0 Iris-setosa
3 4.6 3.1 1.5 0.2 0 Iris-setosa
4 5.0 3.6 1.4 0.2 0 Iris-setosa
In [8]:
df['species'].value_counts()
Out[8]:
Iris-setosa         50
Iris Versicolour    50
Iris Virginica      50
Name: species, dtype: int64
In [9]:
plt.figure(figsize=(7,5))
g=sns.scatterplot(data=df, x="sepal length (cm)", y="sepal width (cm)")
g.set(xlabel ="sepal length (cm) ", ylabel = "sepal width (cm) ", title ='Iris sepal length vs sepal width (cm)')

#plt.xlabel("sepal length (cm) ",fontsize=13)
#plt.ylabel("sepal width (cm) ", fontsize=13)
#plt.title('Iris sepal length vs sepal width (cm)', fontsize=17)
plt.show()
In [10]:
plt.figure(figsize=(7,5))
g=sns.scatterplot(data=df, x="petal length (cm)", y="petal width (cm)")
g.set(xlabel ="petal length (cm)", ylabel = "petal width (cm) ", title ='Iris petal length vs petal width')

plt.show()

iris_setosa=df.loc[df["species"]=="Iris-setosa"] iris_virginica=df.loc[df["species"]=="Iris-virginica"] iris_versicolor=df.loc[df["species"]=="Iris-versicolor"]

Distribution plots¶

plot each flower to a histogram¶

In [11]:
#fig, ax = plt.subplots(figsize=(10, 8))
g=sns.FacetGrid(df, hue="species", height=4)
g.map(sns.histplot,"petal length (cm)", kde=True).add_legend()
h=sns.FacetGrid(df, hue="species", height=4)
h.map(sns.histplot,"petal width (cm)", kde=True).add_legend()
i=sns.FacetGrid(df, hue="species", height=4)
i.map(sns.histplot,"sepal length (cm)", kde=True).add_legend()

plt.show()

Boxplot¶

In [12]:
sns.boxplot(x="species",y="petal length (cm)",data=df)
plt.show()
In [13]:
sns.violinplot(x="species",y="petal width (cm)",data=df)
plt.show()

Correlation¶

In [14]:
df.corr()
Out[14]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) target
sepal length (cm) 1.000000 -0.117570 0.871754 0.817941 0.782561
sepal width (cm) -0.117570 1.000000 -0.428440 -0.366126 -0.426658
petal length (cm) 0.871754 -0.428440 1.000000 0.962865 0.949035
petal width (cm) 0.817941 -0.366126 0.962865 1.000000 0.956547
target 0.782561 -0.426658 0.949035 0.956547 1.000000
In [15]:
sns.set_style("whitegrid")
sns.pairplot(df, hue="species", height=3).add_legend()
plt.show()
In [16]:
x=df[['sepal length (cm)','sepal width (cm)','petal length (cm)','petal width (cm)']]

1. Let's consider all the features¶

In [17]:
input=df[['sepal length (cm)','sepal width (cm)','petal length (cm)','petal width (cm)']]
input.head()
Out[17]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm)
0 5.1 3.5 1.4 0.2
1 4.9 3.0 1.4 0.2
2 4.7 3.2 1.3 0.2
3 4.6 3.1 1.5 0.2
4 5.0 3.6 1.4 0.2

Find optimum number of cluster¶

In [18]:
from sklearn.cluster import KMeans
sse = [] #SUM OF SQUARED ERROR
for k in range(1,11):
    km = KMeans(n_clusters=k, random_state=2)
    km.fit(input)
    sse.append(km.inertia_)
sse
Out[18]:
[681.3706,
 152.3479517603579,
 78.85144142614601,
 57.22847321428571,
 46.44618205128205,
 39.03998724608725,
 34.723277629233515,
 30.06311061745273,
 28.08941131237184,
 25.967077979038507]
In [19]:
sns.set_style("whitegrid")
g=sns.lineplot(x=range(1,11), y=sse)

g.set(xlabel ="Number of cluster (k)", ylabel = "Sum Squared Error", title ='Elbow Method')

plt.show()
In [20]:
kmeans = KMeans(n_clusters = 3, random_state = 2)
pred = kmeans.fit_predict(input)
pred
Out[20]:
array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 2, 2, 2, 2, 1, 2, 2, 2,
       2, 2, 2, 1, 1, 2, 2, 2, 2, 1, 2, 1, 2, 1, 2, 2, 1, 1, 2, 2, 2, 2,
       2, 1, 2, 2, 2, 2, 1, 2, 2, 2, 1, 2, 2, 2, 1, 2, 2, 1])
In [21]:
df['pred']=pred
df.head(3)
Out[21]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) target species pred
0 5.1 3.5 1.4 0.2 0 Iris-setosa 0
1 4.9 3.0 1.4 0.2 0 Iris-setosa 0
2 4.7 3.2 1.3 0.2 0 Iris-setosa 0
In [22]:
df[df.pred==0].head(3)
Out[22]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) target species pred
0 5.1 3.5 1.4 0.2 0 Iris-setosa 0
1 4.9 3.0 1.4 0.2 0 Iris-setosa 0
2 4.7 3.2 1.3 0.2 0 Iris-setosa 0
In [23]:
df[df.pred==1].tail(3)
Out[23]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) target species pred
142 5.8 2.7 5.1 1.9 2 Iris Virginica 1
146 6.3 2.5 5.0 1.9 2 Iris Virginica 1
149 5.9 3.0 5.1 1.8 2 Iris Virginica 1
In [24]:
df[df.pred==2].tail(3)
Out[24]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) target species pred
145 6.7 3.0 5.2 2.3 2 Iris Virginica 2
147 6.5 3.0 5.2 2.0 2 Iris Virginica 2
148 6.2 3.4 5.4 2.3 2 Iris Virginica 2
In [25]:
df.groupby(['target','pred']).count()
Out[25]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) species
target pred
0 0 50 50 50 50 50
1 1 48 48 48 48 48
2 2 2 2 2 2
2 1 14 14 14 14 14
2 36 36 36 36 36
In [26]:
df['pred_species'] = df['pred'].replace(to_replace=[0,1,2], value=['Iris-setosa','Iris Versicolour','Iris Virginica'])
In [27]:
from sklearn.metrics import accuracy_score
accuracy_score(df.species, df.pred_species, normalize=True)
Out[27]:
0.8933333333333333

Accuracy = 89.33%¶

In [28]:
confusion_matrix = pd.crosstab(df['species'], df['pred_species'], rownames=['Actual'], colnames=['Predicted'])

sns.heatmap(confusion_matrix, annot=True)
plt.show()
In [29]:
kmeans.cluster_centers_
Out[29]:
array([[5.006     , 3.428     , 1.462     , 0.246     ],
       [5.9016129 , 2.7483871 , 4.39354839, 1.43387097],
       [6.85      , 3.07368421, 5.74210526, 2.07105263]])
In [30]:
#Actual
plt.figure(figsize=(7,5))
color=['blue','orange','green']
g=sns.scatterplot(data=df, x="petal length (cm)", y="petal width (cm)", hue='species', palette=color, s=75)
g.set(title ='Actual')

plt.show()

#Predicted
plt.figure(figsize=(7,5))
color=['blue','green','orange']
h=sns.scatterplot(data=df, x="petal length (cm)", y="petal width (cm)", hue='pred_species', palette=color, s=75)
sns.scatterplot(x=kmeans.cluster_centers_[:,2],y=kmeans.cluster_centers_[:,3],
                s=450, color='black',marker='*',label='centroid')
h.set(title ='Predicted')
plt.show()
In [31]:
#Actual
plt.figure(figsize=(7,5))
color=['blue','orange','green']
g=sns.scatterplot(data=df,  x="petal length (cm)",y="sepal width (cm)",  hue='species', palette=color, s=75)
g.set(title ='Actual')

plt.show()

#Predicted
plt.figure(figsize=(7,5))
color=['blue','green','orange']
h=sns.scatterplot(data=df,  x="petal length (cm)",y="sepal width (cm)", hue='pred_species', palette=color, s=75)
sns.scatterplot(x=kmeans.cluster_centers_[:,2],y=kmeans.cluster_centers_[:,1],
                s=450, color='black',marker='*',label='centroid')
h.set(title ='Predicted')
plt.show()
In [32]:
#Actual
plt.figure(figsize=(7,5))
color=['blue','orange','green']
g=sns.scatterplot(data=df, x="sepal length (cm)", y="petal width (cm)", hue='species', palette=color, s=75)
g.set( title ='Actual')

plt.show()

#Predicted
plt.figure(figsize=(7,5))
color=['blue','green','orange']
h=sns.scatterplot(data=df, x="sepal length (cm)", y="petal width (cm)", hue='pred_species', palette=color, s=75)
sns.scatterplot(x=kmeans.cluster_centers_[:,0],y=kmeans.cluster_centers_[:,3],
                s=450, color='black',marker='*',label='centroid')
h.set(title ='Predicted')
plt.show()
In [33]:
#Actual
plt.figure(figsize=(7,5))
g=sns.scatterplot(data=df, x="sepal length (cm)", y="sepal width (cm)", hue='species', palette='deep', s=75)

g.set(title ='Actual')

plt.show()


#Predicted
plt.figure(figsize=(7,5))
g=sns.scatterplot(data=df, x="sepal length (cm)", y="sepal width (cm)", hue='pred_species', palette='deep', s=75)
sns.scatterplot(x=kmeans.cluster_centers_[:,0],y=kmeans.cluster_centers_[:,1],
                s=450, color='black',marker='*',label='centroid')
g.set(title ='Predicted')

plt.show()
In [34]:
sns.set_style("whitegrid")
sns.pairplot(df, hue="pred_species", height=3).add_legend()

plt.show()

2. Let's consider only 'sepal length (cm)','petal length (cm)','petal width (cm)'¶

In [35]:
input=df[['sepal length (cm)','petal length (cm)','petal width (cm)']]
input.head()
Out[35]:
sepal length (cm) petal length (cm) petal width (cm)
0 5.1 1.4 0.2
1 4.9 1.4 0.2
2 4.7 1.3 0.2
3 4.6 1.5 0.2
4 5.0 1.4 0.2
In [36]:
kmeans = KMeans(n_clusters = 3, random_state = 2)
pred = kmeans.fit_predict(input)
pred
Out[36]:
array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 2, 2, 2, 2, 1, 2, 2, 2,
       2, 2, 2, 1, 1, 2, 2, 2, 2, 1, 2, 1, 2, 1, 2, 2, 1, 1, 2, 2, 2, 2,
       2, 1, 2, 2, 2, 2, 1, 2, 2, 2, 1, 2, 2, 2, 1, 2, 2, 1])
In [37]:
df['pred']=pred
In [38]:
df.groupby(['target','pred']).count()
Out[38]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) species pred_species
target pred
0 0 50 50 50 50 50 50
1 1 48 48 48 48 48 48
2 2 2 2 2 2 2
2 1 14 14 14 14 14 14
2 36 36 36 36 36 36
In [39]:
df['pred_species1'] = df['pred'].replace(to_replace=[0,1,2], 
                                               value=['Iris-setosa','Iris Versicolour','Iris Virginica'])
In [40]:
accuracy_score(df.species, df.pred_species1, normalize=True)
Out[40]:
0.8933333333333333

accuracy=89.33%¶

In [41]:
confusion_matrix = pd.crosstab(df['species'], df['pred_species1'], rownames=['Actual'], colnames=['Predicted'])

sns.heatmap(confusion_matrix, annot=True)
plt.show()
In [42]:
kmeans.cluster_centers_
Out[42]:
array([[5.006     , 1.462     , 0.246     ],
       [5.9016129 , 4.39354839, 1.43387097],
       [6.85      , 5.74210526, 2.07105263]])
In [43]:
#Actual
plt.figure(figsize=(7,5))
#color=['blue','orange','green']
g=sns.scatterplot(data=df, x="petal length (cm)", y="petal width (cm)", hue='species', palette=color, s=75)
g.set(title ='Actual')

plt.show()

#Predicted
plt.figure(figsize=(7,5))
#color=['blue','green','orange']
h=sns.scatterplot(data=df, x="petal length (cm)", y="petal width (cm)", hue='pred_species1', palette=color, s=75)
sns.scatterplot(x=kmeans.cluster_centers_[:,1],y=kmeans.cluster_centers_[:,2],
                s=450, color='black',marker='*',label='centroid')
h.set(title ='Predicted')
plt.show()
In [44]:
#Actual
plt.figure(figsize=(7,5))
#color=['blue','orange','green']
g=sns.scatterplot(data=df, x="sepal length (cm)", y="petal width (cm)", hue='species', palette=color, s=75)
g.set(title ='Actual')

plt.show()

#Predicted
plt.figure(figsize=(7,5))
#color=['blue','green','orange']
h=sns.scatterplot(data=df, x="sepal length (cm)", y="petal width (cm)", hue='pred_species1', palette=color, s=75)
sns.scatterplot(x=kmeans.cluster_centers_[:,0],y=kmeans.cluster_centers_[:,2],
                s=450, color='black',marker='*',label='centroid')
h.set(title ='Predicted')
plt.show()
In [45]:
#Actual
plt.figure(figsize=(7,5))
#color=['blue','orange','green']
g=sns.scatterplot(data=df, x="sepal length (cm)", y="petal length (cm)", hue='species', palette=color, s=75)
g.set(title ='Actual')

plt.show()

#Predicted
plt.figure(figsize=(7,5))
#color=['blue','green','orange']
h=sns.scatterplot(data=df, x="sepal length (cm)", y="petal length (cm)", hue='pred_species1', palette=color, s=75)
sns.scatterplot(x=kmeans.cluster_centers_[:,0],y=kmeans.cluster_centers_[:,1],
                s=450, color='black',marker='*',label='centroid')
h.set(title ='Predicted')
plt.show()
In [46]:
Target=df[['sepal length (cm)','petal length (cm)','petal width (cm)','target','species'
           ,'pred','pred_species1']]
sns.set_style("whitegrid")
sns.pairplot(Target, hue="pred_species1", height=3).add_legend()
plt.show()

3. Let's consider 'sepal width (cm)','petal length (cm)','petal width (cm)'¶

In [47]:
input=df[['sepal width (cm)','petal length (cm)','petal width (cm)',]]
input.head()
Out[47]:
sepal width (cm) petal length (cm) petal width (cm)
0 3.5 1.4 0.2
1 3.0 1.4 0.2
2 3.2 1.3 0.2
3 3.1 1.5 0.2
4 3.6 1.4 0.2
In [48]:
kmeans = KMeans(n_clusters = 3, random_state = 2)
pred = kmeans.fit_predict(input)
pred
Out[48]:
array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 2, 2, 2, 2, 2, 0, 2, 2, 2,
       2, 2, 2, 2, 2, 2, 2, 2, 2, 0, 2, 2, 2, 0, 2, 2, 0, 2, 2, 2, 2, 2,
       2, 2, 2, 2, 2, 2, 0, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2])
In [49]:
df['pred']=pred
In [50]:
df.groupby(['target','pred']).count()
Out[50]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) species pred_species pred_species1
target pred
0 1 50 50 50 50 50 50 50
1 0 48 48 48 48 48 48 48
2 2 2 2 2 2 2 2
2 0 5 5 5 5 5 5 5
2 45 45 45 45 45 45 45
In [51]:
df['pred_species2'] = df['pred'].replace(to_replace=[1,0,2], 
                                               value=['Iris-setosa','Iris Versicolour','Iris Virginica'])
In [52]:
accuracy_score(df.species, df.pred_species2, normalize=True)
Out[52]:
0.9533333333333334

accuracy=95.33%¶

In [53]:
confusion_matrix = pd.crosstab(df['species'], df['pred_species2'], rownames=['Actual'], colnames=['Predicted'])

sns.heatmap(confusion_matrix, annot=True)
plt.show()
In [54]:
kmeans.cluster_centers_
Out[54]:
array([[2.75471698, 4.28113208, 1.3509434 ],
       [3.428     , 1.462     , 0.246     ],
       [3.00425532, 5.6106383 , 2.04255319]])
In [55]:
#Actual
plt.figure(figsize=(7,5))
#color=['blue','orange','green']
g=sns.scatterplot(data=df,  x="petal length (cm)", y="sepal width (cm)", hue='species', palette=color, s=75)
g.set(title ='Actual')

plt.show()

#Predicted
plt.figure(figsize=(7,5))
#color=['blue','green','orange']
h=sns.scatterplot(data=df, x="petal length (cm)", y="sepal width (cm)", hue='pred_species2', palette=color, s=75)
sns.scatterplot(x=kmeans.cluster_centers_[:,1],y=kmeans.cluster_centers_[:,0],
                s=450, color='black',marker='*',label='centroid')
h.set(title ='Predicted')
plt.show()
In [56]:
#Actual
plt.figure(figsize=(7,5))
#color=['blue','orange','green']
g=sns.scatterplot(data=df,  x="petal width (cm)", y="sepal width (cm)", hue='species', palette=color, s=75)
g.set(title ='Actual')

plt.show()

#Predicted
plt.figure(figsize=(7,5))
#color=['blue','green','orange']
h=sns.scatterplot(data=df, x="petal width (cm)", y="sepal width (cm)", hue='pred_species2', palette=color, s=75)
sns.scatterplot(x=kmeans.cluster_centers_[:,2],y=kmeans.cluster_centers_[:,0],
                s=450, color='black',marker='*',label='centroid')
h.set(title ='Predicted')
plt.show()
In [57]:
#Actual
plt.figure(figsize=(7,5))
#color=['blue','orange','green']
g=sns.scatterplot(data=df,  x="petal length (cm)", y="petal width (cm)", hue='species', palette=color, s=75)
g.set(title ='Actual')

plt.show()

#Predicted
plt.figure(figsize=(7,5))
#color=['blue','green','orange']
h=sns.scatterplot(data=df, x="petal length (cm)", y="petal width (cm)", hue='pred_species2', palette=color, s=75)
sns.scatterplot(x=kmeans.cluster_centers_[:,1],y=kmeans.cluster_centers_[:,2],
                s=450, color='black',marker='*',label='centroid')
h.set(title ='Predicted')
plt.show()
In [58]:
Target=df[['sepal width (cm)','petal length (cm)','petal width (cm)','target','species',
           'pred','pred_species2']]
sns.set_style("whitegrid")
sns.pairplot(Target, hue="pred_species2", height=3).add_legend()
plt.show()

4. Lets consider "petal length" and "petal width"¶

In [59]:
Input= df[['petal length (cm)','petal width (cm)']]
Input.head()
Out[59]:
petal length (cm) petal width (cm)
0 1.4 0.2
1 1.4 0.2
2 1.3 0.2
3 1.5 0.2
4 1.4 0.2
In [60]:
kmeans = KMeans(n_clusters=3, random_state=2)
pred = kmeans.fit_predict(Input)
pred
Out[60]:
array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
       2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 2, 2, 2, 2, 2, 1, 2, 2, 2, 2,
       2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1])
In [61]:
df['pred']=pred
In [62]:
df.groupby(['target','pred']).count()
Out[62]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) species pred_species pred_species1 pred_species2
target pred
0 0 50 50 50 50 50 50 50 50
1 1 2 2 2 2 2 2 2 2
2 48 48 48 48 48 48 48 48
2 1 46 46 46 46 46 46 46 46
2 4 4 4 4 4 4 4 4
In [63]:
df['pred_species3'] = df['pred'].replace(to_replace=[0,2,1], 
                                               value=['Iris-setosa','Iris Versicolour','Iris Virginica'])
In [64]:
accuracy_score(df.species, df.pred_species3, normalize=True)
Out[64]:
0.96

accuracy = 96%¶

In [65]:
confusion_matrix = pd.crosstab(df['species'], df['pred_species3'], rownames=['Actual'], colnames=['Predicted'])

sns.heatmap(confusion_matrix, annot=True)
plt.show()
In [66]:
kmeans.cluster_centers_
Out[66]:
array([[1.462     , 0.246     ],
       [5.59583333, 2.0375    ],
       [4.26923077, 1.34230769]])
In [67]:
#Actual
plt.figure(figsize=(7,5))
#color=['blue','orange','green']
g=sns.scatterplot(data=df,  x="petal length (cm)", y="petal width (cm)", hue='species', palette=color, s=75)
g.set(title ='Actual')

plt.show()

#Predicted
plt.figure(figsize=(7,5))
#color=['blue','green','orange']
h=sns.scatterplot(data=df, x="petal length (cm)", y="petal width (cm)", hue='pred_species3', palette=color, s=75)
sns.scatterplot(x=kmeans.cluster_centers_[:,0],y=kmeans.cluster_centers_[:,1],
                s=450, color='black',marker='*',label='centroid')
h.set(title ='Predicted')
plt.show()
In [68]:
Target=df[["petal length (cm)","petal width (cm)",'target','species',
           'pred', 'pred_species3']]
sns.set_style("whitegrid")
sns.pairplot(Target, hue="pred_species3", height=3).add_legend()
plt.show()

Preprocessing using min max scaler¶

image.png

In [69]:
from sklearn.decomposition import NMF
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import MinMaxScaler
import warnings
warnings.filterwarnings("ignore")
In [70]:
pipe = Pipeline(steps=[
    ('scaler', MinMaxScaler()),
    ('pca',  NMF(n_components=2, init='random', random_state=2)),     
    ('KMeans', KMeans(n_clusters=3, random_state=2)),
])

Considering all four features¶

In [71]:
x=df[['sepal length (cm)','sepal width (cm)','petal length (cm)','petal width (cm)']]
In [72]:
pipe.fit(x)
Out[72]:
Pipeline(steps=[('scaler', MinMaxScaler()),
                ('pca', NMF(init='random', n_components=2, random_state=2)),
                ('KMeans', KMeans(n_clusters=3, random_state=2))])
In [73]:
df['cluster']=pipe.predict(x)
df.groupby(['target','cluster']).count()
Out[73]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) species pred pred_species pred_species1 pred_species2 pred_species3
target cluster
0 1 50 50 50 50 50 50 50 50 50 50
1 0 48 48 48 48 48 48 48 48 48 48
2 2 2 2 2 2 2 2 2 2 2
2 0 14 14 14 14 14 14 14 14 14 14
2 36 36 36 36 36 36 36 36 36 36
In [74]:
df['cluster'] = df['cluster'].replace(to_replace=[1,0,2], 
                                               value=['Iris-setosa','Iris Versicolour','Iris Virginica'])
In [75]:
accuracy_score(df.species,df.cluster,normalize=True)
Out[75]:
0.8933333333333333

accuracy = 89.33%¶

In [76]:
pipe.predict([[6,2,5,2]])
Out[76]:
array([0])
In [77]:
kmeans.predict([[5.1,1.8]])
Out[77]:
array([1])

Let's consider only ['sepal length (cm)','petal length (cm)','petal width (cm)']¶

In [78]:
x=df[['sepal length (cm)','petal length (cm)','petal width (cm)']]
In [79]:
pipe.fit(x)
df['pred']=pipe.predict(x)
df.groupby(['target','pred']).count()
Out[79]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) species pred_species pred_species1 pred_species2 pred_species3 cluster
target pred
0 0 50 50 50 50 50 50 50 50 50 50
1 1 2 2 2 2 2 2 2 2 2 2
2 48 48 48 48 48 48 48 48 48 48
2 1 46 46 46 46 46 46 46 46 46 46
2 4 4 4 4 4 4 4 4 4 4
In [80]:
df['cluster'] = df['pred'].replace(to_replace=[0,2,1], 
                                               value=['Iris-setosa','Iris Versicolour','Iris Virginica'])
accuracy_score(df.species,df.cluster,normalize=True)
Out[80]:
0.96

accuracy=96%¶

In [81]:
sns.set_style("whitegrid")
sns.pairplot(df, hue="cluster", height=3)
plt.show()
In [82]:
pipe.predict([[6, 5, 2]])
Out[82]:
array([1])

Let's consider only ['sepal width (cm)','petal length (cm)','petal width (cm)']¶

In [83]:
x=df[['sepal width (cm)','petal length (cm)','petal width (cm)']]
In [84]:
pipe.fit(x)
df['pred']=pipe.predict(x)
df.groupby(['target','pred']).count()
Out[84]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) species pred_species pred_species1 pred_species2 pred_species3 cluster
target pred
0 0 50 50 50 50 50 50 50 50 50 50
1 1 2 2 2 2 2 2 2 2 2 2
2 48 48 48 48 48 48 48 48 48 48
2 1 46 46 46 46 46 46 46 46 46 46
2 4 4 4 4 4 4 4 4 4 4
In [85]:
df['cluster'] = df['pred'].replace(to_replace=[0,1,2], 
                                               value=['Iris-setosa','Iris Versicolour','Iris Virginica'])
accuracy_score(df.species,df.cluster,normalize=True)
Out[85]:
0.37333333333333335

Accuracy =96%¶

In [86]:
pipe.predict([[6, 5, 2]])
Out[86]:
array([1])

Let's consider only ['petal length (cm)','petal width (cm)']¶

In [87]:
x=df[['petal length (cm)','petal width (cm)']]
In [88]:
pipe = Pipeline(steps=[
    ('scaler', MinMaxScaler()),
    #('pca',  NMF(n_components=2, init='random', random_state=2)),     
    ('KMeans', KMeans(n_clusters=3, random_state=2)),
])
In [89]:
pipe.fit(x)
df['pred']=pipe.predict(x)
df.groupby(['target','pred']).count()
Out[89]:
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) species pred_species pred_species1 pred_species2 pred_species3 cluster
target pred
0 0 50 50 50 50 50 50 50 50 50 50
1 1 48 48 48 48 48 48 48 48 48 48
2 2 2 2 2 2 2 2 2 2 2
2 1 4 4 4 4 4 4 4 4 4 4
2 46 46 46 46 46 46 46 46 46 46
In [90]:
df['cluster'] = df['pred'].replace(to_replace=[0,1,2], 
                                               value=['Iris-setosa','Iris Versicolour','Iris Virginica'])
accuracy_score(df.species,df.cluster,normalize=True)
Out[90]:
0.96

Accuracy = 96%¶

In [91]:
pipe.predict([[5,2]])
Out[91]:
array([2])
In [92]:
kmeans.predict([[5,2]])
Out[92]:
array([1])

Actually for kmeans 1='Iris Virginica', 2= 'Iris Versicolour'and 0='Iris-setosa' while for pipe 2='Iris Virginica', 1= 'Iris Versicolour'and 0='Iris-setosa'

  • Both the predictions are same
In [93]:
pipe.predict([[3,2]])
Out[93]:
array([1])
In [94]:
kmeans.predict([[3,2]])
Out[94]:
array([2])
In [95]:
pipe.predict([[2,2]])
Out[95]:
array([1])
In [96]:
kmeans.predict([[2,2]])
Out[96]:
array([0])
In [97]:
pipe.predict([[1,1]])
Out[97]:
array([0])
In [98]:
kmeans.predict([[1,1]])
Out[98]:
array([0])
In [99]:
kmeans.cluster_centers_
Out[99]:
array([[1.462     , 0.246     ],
       [5.59583333, 2.0375    ],
       [4.26923077, 1.34230769]])
In [100]:
#Actual
plt.figure(figsize=(7,5))
#color=['blue','orange','green']
g=sns.scatterplot(data=df,  x="petal length (cm)", y="petal width (cm)", hue='species', palette=color, s=75)
g.set(title ='Actual')

plt.show()

#Predicted
plt.figure(figsize=(7,5))
#color=['blue','green','orange']
h=sns.scatterplot(data=df, x="petal length (cm)", y="petal width (cm)", hue='cluster', palette=color, s=75)
sns.scatterplot(x=kmeans.cluster_centers_[:,0],y=kmeans.cluster_centers_[:,1],
                s=450, color='black',marker='*',label='centroid')
h.set(title ='Predicted')
plt.show()

Final Accuracy Score= 96%¶